Movalytics

Home
/ Movalytics

Live product

Clinical motion analysis from a single smartphone recording.

See a real analysis

Three recordings, four views. Switch instantly.

Recording

View

Live output from app.babon.eu, not pre-rendered.Open in app.babon.eu →

Methode

The method

Everyone else who tries to extract clinical joint angles from video today — Stanford OpenCap, Theia3D, BioPose, the recently released OpenCap Monocular — builds an intermediate skeleton model and runs an inverse-kinematics solver on top of it. Hundreds to thousands of lines of pipeline code, usually on OpenSim or MuJoCo, seconds to minutes of compute per recording.

Parametric body models actually compute the rotation matrices you need for those clinical angles already; nobody just uses them directly. We do. We decompose them and calibrate against ISB convention in place. That saves roughly an order of magnitude in code and three orders of magnitude in compute, while working off a single smartphone recording where Stanford needs two cameras. The patent on this approach was filed in April 2026 and covers all parametric body models, not just the ones we currently use.

	Rest of the field	Movalytics
Pipeline	parametric model → skeleton → IK solver → ISB angles	parametric model → ISB angles direct
Pipeline complexity	~1700 lines, OpenSim / MuJoCo	~200 lines
Compute per recording	seconds to minutes	milliseconds
Cameras	multi-view or marker-based lab	single smartphone

How it works

Three steps, no manual cleanup.

Film

A smartphone recording from the side or the front, depending on which planes you want to see. No markers on the body, no calibration frames.

Upload

The video goes to app.babon.eu. We detect the person, fit a parametric body model and pull the ISB angles directly out of it. Minutes.

Report

You get joint angles per plane, gait parameters, ground reaction forces and symmetry indicators back, plus CSV and JSON to keep working in your own tools.

Three ways to use Movalytics

Physiotherapy and research are free in exchange for the right to use the anonymised data. Education is a separate service with its own pricing because education recordings come with different quality guarantees and do not belong in the research database.

Physiotherapy

Objective documentation in daily practice

You upload a gait video of a patient and get a clinical report back within minutes — joint angles, gait parameters, symmetry. That is an objective measurement alongside your clinical judgement, usable in the patient record and in the conversation with the patient. Free for practices that join the database: we keep only the movement parameters (no video, no patient data) and use those to build normative references — references that in turn make the tool better for everyone.

Request early access

Research

Free analysis, in exchange for data and co-publication

If you have video data from a specific research project — back pain, post-surgical recovery, sports, elderly care — we analyse it for free, including support during processing. In exchange we get to use the anonymised kinematics and we publish together. You keep control over your protocol and over what happens with your data. Serious graduate projects count as research too.

Discuss a research partnership

Education

Learn biomechanics by measuring it yourself

For movement-science and movement-health programmes. Students record themselves and each other, upload the videos, and see the movement come back as joint-angle charts and gait parameters — comparing the theory from lectures with the reality of their own gait. Honestly: this mode is still batch-only right now, so what you upload today you get back tomorrow. Fine for classroom use, not for the clinical flow. Pricing and setup are agreed per programme.

Request a conversation

What you get back

Joint angles

Hip, knee, ankle, shoulder, elbow, wrist, neck, trunk per plane (sagittal, frontal, transversal). ISB convention.

Gait parameters

Velocity, cadence, stride length, double/single support phase, symmetry index. Per cycle and averaged.

Forces

Ground reaction forces, peak vGRF, loading rate, push-off. Embedded GaitDynamics integration (Stanford).

L/R symmetry

Symmetry index per parameter with colour-coded severity. See immediately where it is off.

Export

CSV per frame, JSON per recording, PNG charts. Compatible with Excel, Python, R, MATLAB.

3D viewer

Interactive 3D skeleton in the browser, synced with video and angle charts. No plug-ins, no downloads.

Validation

We have compared Movalytics against marker-based motion capture on the Stanford LabValidation dataset, and a collaboration with Jaap Jansen and his team at Hogeschool Utrecht is under way to repeat that comparison on their own lab data. The full numbers are going into a peer-reviewed paper currently in preparation; we will link it here the moment it is published.

Kardolus, Hendriks, Jansen — in preparation.

Regulatory

MDR Class I — self-declaration in progress

Movalytics is categorised as an MDR Class I medical device. That is the lightest class, for measurement instruments that report kinematic parameters without issuing a diagnosis or treatment recommendation. The self-declaration is in progress; we expect CE marking during 2026. Clinical judgement stays with the clinician.

Interested in early access?

Get in touch